The Curious Case of @Fascist Kirk - Or, Why I Hate Stickers and People Who Care About Them

You'll notice it if you try to phonepost on 4cuck while using mobile data. I'd say like 30% of the time, you'll end up intercepting some ancient ban or warning that got handed out to some fag who used that IPv6 address years ago.
To be fair you'd have to be an idiot to use 4chan any time in the past more or less 2 decades but I'll take your word for it
 
I'd be tempted to plug an LLM into the logs to analyze user behavior and detect socks. Should be simple enough to feed logs to some lightweight model like Qwen2-7B-Instruct.

Bonus points if the LLM can post on its own and starts calling out users for being socks.
how is it gonna be able to differentiate between users when everybody starts saying NIGGERFAGGOT TRANNYNIGGER NIGGERPUSSY as a kind of anti-bot jamming? You can circumvent no-no-word restrictions, of course, but it still seems like a huge issue. Autists like the average KF user would probably (:optimistic:) be the type to devise some form of countermeasure or encryption that reduces the efficacy of the program.
 
  • Agree
Reactions: Fascist Kirk
VBHN.gif
 
how is it gonna be able to differentiate between users when everybody starts saying NIGGERFAGGOT TRANNYNIGGER NIGGERPUSSY as a kind of anti-bot jamming? You can circumvent no-no-word restrictions, of course, but it still seems like a huge issue. Autists like the average KF user would probably (:optimistic:) be the type to devise some form of countermeasure or encryption that reduces the efficacy of the program.
That's not what it's meant to do, it won't be reading everyone's posts but analyzing logs on the backend.

Null outlined all the steps he took in the first post, all of those steps can be done by AI with access to backend logs.

Checking IP collisions? Easy. Identifying sticker spergouts? Easy. Then just try matching patterns, check for low post count users who heavily use stickers, note their targets, time of their posts, etc.
 
Back